© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 33-GEMMASCOPE-RES-16K
    6. 319
    Prev
    Next
    INDEX
    Explanations

    programming loops and control structures, particularly 'for' and 'while' statements

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 33-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_33/width_16k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.33.hook_resid_post
    Hook Layer
    33
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    BufferException
    -0.62
     насељу
    -0.62
    principalColumn
    -0.60
    SharedDtor
    -0.59
     Administrativna
    -0.57
    GOTREF
    -0.54
    ArrowToggle
    -0.53
    Personendaten
    -0.53
    PROMOTED
    -0.52
     تانيه
    -0.52
    POSITIVE LOGITS
    each
    0.43
    Each
    0.39
     each
    0.37
     Each
    0.36
    ENSIS
    0.35
     EACH
    0.34
    EACH
    0.34
    foreach
    0.34
     chaque
    0.33
    各
    0.33
    Activations Density 0.077%

    No Known Activations